This study examines how firms can optimize Referral Reward Programs (RRPs) by addressing two conflicting incentives: referrers' social concerns and recipients' psychological reactance. It investigates the optimal RRP design under these tensions and assesses whether market segmentation enhances profitability.
A nested Stackelberg model is constructed to analyze the behavior between the two-level referrers, customers, and firms. A theoretical comparison of a firm's profit has been made when implementing mass market advertising and RRP.
(1) The social relationship among customers has economic value. (2) Monetary rewards for referrers should be differentiated across platforms with varying social tie strength. (3) Firms need to appropriately incentivize those referrers who are worried about their perceived image during product recommendations. (4) RRPs should not be used in markets that are already efficient at attracting customers or when the latter have weak social connections online.
Two limitations concern the exogenous treatment of the Incumbent–Customer Ratio and the restriction to single-receiver referrals. Future research should develop endogenous models and incorporate multi-receiver frameworks.
Firms should align RRPs with customers' social ties and brand attributes, avoiding uniform rewards. Strong-tie referrals (e.g. friends) require higher monetary incentives to offset social concerns, whereas moderate-to-weak ties (e.g. acquaintances) are better served by lower rewards to reduce reactance.
Social ties have evolved into measurable economic assets in marketing. Firms must balance leveraging social media ecosystems with their preservation, as undermining trust renders social marketing ineffective.
This study develops a novel multilevel referrer framework (first- and second-level) that better reflects real-world RRP dynamics. It incorporates both social image concern and psychological reactance into incentive design, and provides clear managerial guidelines for choosing between mass advertising and RRPs based on decision thresholds.
